ERROR: Unable to locate IOAPIC for GSI

raggerv8 · 2016-10-21 08:07:59

Today, using the vanilla kernel (4.8.2-1), I am noticing this error in the console and journalctl:

Oct 21 09:59:10 noname kernel: ERROR: Unable to locate IOAPIC for GSI 37
Oct 21 09:59:10 noname kernel: snd_hda_intel 0000:00:03.0: HSW/BDW HD-audio HDMI/DP requires binding with gfx driver

I include both errors, but the interesting one is the top one called ERROR: Unable to locate IOAPIC for GSI. Everything seems to work fine, but why do I get that? It seems like the first time I got it was yesterday morning:

Oct 20 07:54:00 noname kernel: ERROR: Unable to locate IOAPIC for GSI 37

If you compare the above with my pacman log, you can see that I updated my system, linux-ck which I was using updated, so I rebooted. This was the first time I got this issue, but as I said, this error shows up even using the vanilla kernel, which did not update.

I've googled, but I only find really old issues and there seems to be no explanation or fix.

Edit: The cause of this error is Linux 4.8 on Haswell CPU's (it seems).

Last edited by raggerv8 (2016-10-23 18:47:49)

mrlamud · 2016-10-21 11:15:34

I do also confirm this same error in journal since kernel 4.8 but system runs fine.

Some details which might be helpful:-

lamud@archbox:~/ > uname -a
Linux archbox 4.8.3-1-ARCH #1 SMP PREEMPT Fri Oct 21 07:10:48 ICT 2016 x86_64 GNU/Linux
lamud@archbox:~/ > journalctl -xb | grep ERROR
Oct 21 15:08:50 archbox kernel: ERROR: Unable to locate IOAPIC for GSI 37
lamud@archbox:~/ > lspci -tv
-[0000:00]-+-00.0  Intel Corporation 4th Gen Core Processor DRAM Controller
           +-02.0  Intel Corporation Xeon E3-1200 v3/4th Gen Core Processor Integrated Graphics Controller
           +-03.0  Intel Corporation Xeon E3-1200 v3/4th Gen Core Processor HD Audio Controller
           +-14.0  Intel Corporation 9 Series Chipset Family USB xHCI Controller
           +-16.0  Intel Corporation 9 Series Chipset Family ME Interface #1
           +-19.0  Intel Corporation Ethernet Connection I217-V
           +-1b.0  Intel Corporation 9 Series Chipset Family HD Audio Controller
           +-1c.0-[01]--
           +-1c.3-[02-03]----00.0-[03]--
           +-1f.0  Intel Corporation 9 Series Chipset Family H97 Controller
           +-1f.2  Intel Corporation 9 Series Chipset Family SATA Controller [AHCI Mode]
           \-1f.3  Intel Corporation 9 Series Chipset Family SMBus Controller

Malikos · 2016-10-21 16:17:22

I also have linux-ck and the same error:

 Oct 20 07:54:00 noname kernel: ERROR: Unable to locate IOAPIC for GSI 37

It shows just before the login screen and system won't boot. I have to use linux-zen kernel in order to use my computer at all.

This error startet to show when I updated linux-ck from 4.7.* to 4.8.2. Today it updated to 4.8.3 but the same error still persists.

Does anybody know how to help? Or I need to just wait for more updated linux-ck?

Last edited by Malikos (2016-10-21 16:18:13)

raggerv8 · 2016-10-21 16:23:48

Malikos wrote:

This error startet to show when I updated linux-ck from 4.7.* to 4.8.2. Today it updated to 4.8.3 but the same error still persists.

I suppose it is possible that this issue started with the vanilla 4.8.x kernel, but when I updated to the 4.8.x kernel, I was using linux-ck all the time, so I never noticed it.

rakmial · 2016-10-22 20:28:40

Hey folks,

First time Archer and first post. I caught this same error after the kernel update. The following links give a little more information about the nature of I/O APIC and Global System Interrupts (GSIs), but it's admittedly a bit above my linux paygrade. It seems to be an Intel I/O system. I'll be working on figuring out a solution, but it hasn't adversely affected my OS performance so far - it could be that the PCI in question is not in use on my motherboard.

Arch Screenfetch

moderator edit: Welcome to the forums. The image is too large. I have replaced it with a link for you this time. Please read Code of conduct: Pasting pictures and code and screenshot posting rules. Thanks. --fsckd

Last edited by fsckd (2016-10-23 19:42:51)

agm28011997 · 2016-10-23 18:35:43

in my pc with haswell since the linux update I am receiving this error too... I don't know why but in linux 4.7 not errors..

Malikos · 2016-10-23 18:41:16

agm28011997 wrote:

in my pc with haswell since the linux update I am receiving this error too... I don't know why but in linux 4.7 not errors..

Mine is Haswell as well.
Strange thing is that linux-ck won't boot since 4.8 with the following error, while linux-zen 4.8.3 also shows this error but it boots without problems.

raggerv8 · 2016-10-23 18:45:01

agm28011997 wrote:

in my pc with haswell since the linux update I am receiving this error too... I don't know why but in linux 4.7 not errors..

This makes sense, because I decided to look in my pacman.log for when I upgraded to 4.8.

As I say in the OP, I got this error the first time Oct 20 07:54:00, and I can see that I upgraded from 4.7 to 4.8 a minute before that, and right after I upgraded, I rebooted.

[2016-10-20 07:53] [ALPM] upgraded linux-ck-haswell (4.7.8-1 -> 4.8.2-1)

Last edited by raggerv8 (2016-10-23 18:45:16)

loqs · 2016-10-23 19:35:04

https://git.kernel.org/cgit/linux/kerne … .8.4#n2669 is the producer of error message itself.
Edit:
This seems to be the issue to me https://marc.info/?l=linux-kernel&m=147438001200798&w=2

Last edited by loqs (2016-10-23 20:15:15)

Antares · 2016-10-23 20:41:41

While working in a productive environment on just one machine with ArchLinux here and fearing of crashing the system, may I ask if anybody perhaps tried in a VM to add

noapic

in the kernel line of grub.cfg, recompiled the kernel and rebooted? This was supposed to remove at least the abovementioned error message according to the entries from a SliTaz forum 4 years ago, where a user reported the same problems.

Cotton · 2016-10-23 21:50:27

Still reported at boot using latest kernel package 4.8.4-1-ARCH.
Doesn't show any apparent problems on my system but for those who can't boot, what's the fix?

Last edited by Cotton (2016-10-23 21:50:56)

loqs · 2016-10-24 00:29:53

https://git.kernel.org/cgit/linux/kerne … e6b492171c
https://git.kernel.org/cgit/linux/kerne … 68a533f043
https://git.kernel.org/cgit/linux/kerne … 973ecd63cb

Cotton if you want something to try apply the above three patches in order to the 4.8.4 kernel final patch is the actual fix relies on the previous two.
This is very rushed but will be unable to help for at least a day so putting what I have so far here for others to continue.

agm28011997 · 2016-10-25 00:04:22

Malikos wrote:

agm28011997 wrote:
in my pc with haswell since the linux update I am receiving this error too... I don't know why but in linux 4.7 not errors..
Mine is Haswell as well.
Strange thing is that linux-ck won't boot since 4.8 with the following error, while linux-zen 4.8.3 also shows this error but it boots without problems.

for me the kernel boots too, but the linux ck with bfs, the muqss is not tested by me yet

corner578 · 2016-10-26 18:57:37

I have this situation on AsRock H97 Pro4 also.
But i found that if in /proc/interrupts we see like

cat /proc/interrupts 
           CPU0       CPU1       CPU2       CPU3       CPU4       CPU5       CPU6       CPU7       
  0:         19          0          0          0          0          0          0          0   IO-APIC   2-edge      timer
  1:       2484          0          0          0          0          0          0          0   IO-APIC   1-edge      i8042
  8:          1          0          0          0          0          0          0          0   IO-APIC   8-edge      rtc0
  9:          4          0          0          0          0          0          0          0   IO-APIC   9-fasteoi   acpi
 16:         29          0          0          0          0          0          0          0   IO-APIC  16-fasteoi   ehci_hcd:usb1
 18:          0          0          0          0          0          0          0          0   IO-APIC  18-fasteoi   i801_smbus
 23:         35          0          0          0          0          0          0          0   IO-APIC  23-fasteoi   ehci_hcd:usb2
...

Then all ok (of course if all works normal)

This message is just garbage. I read this in one of letters of Linus Torvalds (don't have and found now)

loqs · 2016-10-26 19:10:21

corner578 wrote:

This message is just garbage. I read this in one of letters of Linus Torvalds (don't have and found now)

So you disagree with the assessment in second link in post #9?
It also does not explain why this might cause some systems not to boot.
But as none of those affected have responded to posts #9 or #12 its hard to say if that is the actual cause.
As I do have an affected system I can not test that myself.
Edit:
corrected reference to post#11 when it should have been post #12

Last edited by loqs (2016-10-26 19:11:38)

corner578 · 2016-10-26 20:30:45

I'm not a guru in linux kernel, but following the next documents I think this is critical message
Linux doc
IOAPIC wiki

I seen that the main is

[    0.000000] IOAPIC[0]: apic_id 8, version 32, address 0xfec00000, GSI 0-23
[    0.838763] ACPI: Using IOAPIC for interrupt routing

My machine work normal, though I have many services for development (such as MariaDB, Redis, local servers, websocket servers etc.) and any disks SSD including.
All works and I have not see any troubles.
Then I think this message is result of incorrect code (maybe) and this message is just garbage.
If you wondering, you can try to create bug maybe. And, I hope, this behavior will fixed in near future

P.S. on my Thinkpad with Intel(R) Core(TM) i7-4702MQ I have not this message

P.S.S. See on desktop zero addresses /proc/iomem - don't know why...
Maybe will try lts kernel later

loqs · 2016-10-26 20:46:32

corner578 wrote:

If you wondering, you can try to create bug maybe.

loqs wrote:

As I do have an affected system I can not test that myself.

Would it not be better for someone affected by the issue to file the bug report?

corner578 wrote:

And, I hope, this behavior will fixed in near future

Did you read the 3rd link in post #12?

corner578 · 2016-10-26 21:29:53

Yes, I do. I read 3rd link

I read also arch/x86/kernel/apic/io_apic.c +2669 now:)
And, by my mind, this message isn't critical

Of course, I'm not a guru in kernel

loqs · 2016-10-26 21:51:57

loqs wrote:

corner578 wrote:
And, I hope, this behavior will fixed in near future
Did you read the 3rd link in post #12?

I asked if you read the link because you seemed to be implying there was no fix for issue.
Yet the commit comes from linus's tree references sysmptoms such as

  [   10.409490] ERROR: Unable to locate IOAPIC for GSI 37

and references https://git.kernel.org/cgit/linux/kerne … 34b59bb5ac
as the causal commit which is occurred during the 4.8 cycle so that matches with the other noted symptom that the issue was not present in 4.7 kernels only 4.8 kernels.
I agree that if all those experiencing the issue choose not to test proposed fixes then there is no way to know for certain if a given solution fixes a given issue.
I agree also it may be unrelated to the none boot issue if one issue is fixed but the other remains that wold demonstrate there is a separate cause which is another reason I requested those affected test the patches.
Edit:

corner578 wrote:

Of course, I'm not a guru in kernel

I am certainly not one either.

Last edited by loqs (2016-10-26 22:07:30)

corner578 · 2016-10-27 15:29:43

Continue...
I have installed linux-lts and have the next result.
This message does not appear

dmesg | grep -i IOAPIC
[    0.000000] IOAPIC[0]: apic_id 8, version 32, address 0xfec00000, GSI 0-23
[    0.245025] ACPI: Using IOAPIC for interrupt routing

and, this is important, proc/iomem have non-zero addresses

cat /proc/iomem
00000000-00000fff : reserved
00001000-0009d7ff : System RAM
0009d800-0009ffff : reserved
000a0000-000bffff : PCI Bus 0000:00
000c0000-000cfdff : Video ROM
000d0000-000d3fff : PCI Bus 0000:00
000d4000-000d7fff : PCI Bus 0000:00
000d8000-000dbfff : PCI Bus 0000:00
000dc000-000dffff : PCI Bus 0000:00
000e0000-000fffff : reserved
...

Then we have a bug yet

loqs · 2016-10-27 16:06:19

corner578 wrote:

and, this is important, proc/iomem have non-zero addresses

corner578 wrote:

Then we have a bug yet

linux-lts is 4.4.27-1
so it would not contain https://git.kernel.org/cgit/linux/kerne … 1e8171dee4
The data being zero means it has been redacted as per that patch.

corner578 · 2016-10-27 16:54:15

I'm totally confused
Is this a bug or not a bug ( Shakespeare's famous question )

I will use linux-lts until

agm28011997 · 2016-11-17 08:40:08

any news? i am upgrading the kernel but the problem is not fixed.... in 4.9 may be?

loqs · 2016-11-17 17:05:39

agm28011997 wrote:

in 4.9 may be?

The three patches I referenced in post #12 are all from the 4.9 series but no one affected by the issue has reported testing them so can not say for certain.

agm28011997 · 2017-03-08 22:39:07

fixed

Arch Linux

#1 2016-10-21 08:07:59

ERROR: Unable to locate IOAPIC for GSI

#2 2016-10-21 11:15:34

Re: ERROR: Unable to locate IOAPIC for GSI

#3 2016-10-21 16:17:22

Re: ERROR: Unable to locate IOAPIC for GSI

#4 2016-10-21 16:23:48

Re: ERROR: Unable to locate IOAPIC for GSI

#5 2016-10-22 20:28:40

Re: ERROR: Unable to locate IOAPIC for GSI

#6 2016-10-23 18:35:43

Re: ERROR: Unable to locate IOAPIC for GSI

#7 2016-10-23 18:41:16

Re: ERROR: Unable to locate IOAPIC for GSI

#8 2016-10-23 18:45:01

Re: ERROR: Unable to locate IOAPIC for GSI

#9 2016-10-23 19:35:04

Re: ERROR: Unable to locate IOAPIC for GSI

#10 2016-10-23 20:41:41

Re: ERROR: Unable to locate IOAPIC for GSI

#11 2016-10-23 21:50:27

Re: ERROR: Unable to locate IOAPIC for GSI

#12 2016-10-24 00:29:53

Re: ERROR: Unable to locate IOAPIC for GSI

#13 2016-10-25 00:04:22

Re: ERROR: Unable to locate IOAPIC for GSI

#14 2016-10-26 18:57:37

Re: ERROR: Unable to locate IOAPIC for GSI

#15 2016-10-26 19:10:21

Re: ERROR: Unable to locate IOAPIC for GSI

#16 2016-10-26 20:30:45

Re: ERROR: Unable to locate IOAPIC for GSI

#17 2016-10-26 20:46:32

Re: ERROR: Unable to locate IOAPIC for GSI

#18 2016-10-26 21:29:53

Re: ERROR: Unable to locate IOAPIC for GSI

#19 2016-10-26 21:51:57

Re: ERROR: Unable to locate IOAPIC for GSI

#20 2016-10-27 15:29:43

Re: ERROR: Unable to locate IOAPIC for GSI

#21 2016-10-27 16:06:19

Re: ERROR: Unable to locate IOAPIC for GSI

#22 2016-10-27 16:54:15

Re: ERROR: Unable to locate IOAPIC for GSI

#23 2016-11-17 08:40:08

Re: ERROR: Unable to locate IOAPIC for GSI

#24 2016-11-17 17:05:39

Re: ERROR: Unable to locate IOAPIC for GSI

#25 2017-03-08 22:39:07

Re: ERROR: Unable to locate IOAPIC for GSI

Board footer